Hi everyone, welcome back to the Harbinger VC podcast. Today we have joining us Peter Cheng, who is founding partner at Eminence Ventures. Eminence just raised $120m for its second fund focused on backing Chinese enterprise software and cloud start-ups. After a decade of torrid start-up growth driven by mobile and consumer internet, China’s enterprise software market is poised for a breakout and expected to reach $40bn in revenue in 2025. Peter shares with us his formative experiences across eBay, AdChina (adtech start-up he founded that was acquired by Alibaba), and Tencent (as GM of ads products), before launching his career as a venture capitalist. Peter provides an overview of China’s cloud market, key opportunities across horizontal and vertical software, AI, and IT infrastructure. He describes key success factors and go-to-market strategy for such companies, including for portfolio companies Tsign and Recurrent.ai. Finally, Peter illustrates key differences between the US and Chinese environments and how that impacts the development trajectory of software companies.
[Editor's note: this interview has been edited andcondensed for clarity. The opinions expressed in this article are Peter’s own and do not reflect the views of Eminence Ventures]
AdChina and eBay
Thanks for joining us today Peter. Tell us a bit more about yourself first, where you're from and what are some of your formative experiences?
Peter: I was born and raised in Taiwan. I went to college at Jiaotong University in Hsinchu, where I was trained as a computer engineer. Post college, after serving two years in the army, I went to the US for my MBA and attended Indiana University. After the MBA, I stayed in the US for about 14 years working for three large companies – FedEx, Oracle, and eBay – as well as a couple of start-ups in Silicon Valley.
In 2007, I moved to Shanghai and co-founded a company called AdChina, which was eventually acquired by Alibaba. After that, I joined Tencent as a General Manager of its advertising platform. And after Tencent, I started my VC career.
Tell us more about your experiences while at eBay.
Peter: I was at eBay for more than four years. Most of the time I was in San Jose, the headquarters. During my last six months, I was actually sent over to Shanghai to help a company that eBay acquired called EachNet. So I was on a short term assignment helping them with product marketing and business development. That was pretty much my first taste of China.
During that six-month period, I saw first-hand the fierce competition and the battle against Alibaba's Taobao. One thing I learned was that the Western mindset often doesn’t work in China. For example, eBay made a lot of acquisitions previously in Europe, France, Germany, and quickly moved them to the eBay main platform. When eBay acquired EachNet, they moved it to the global platform as well.
When you do that, you slow down the development of local features. And when you are trying to compete against Alibaba and you launch features much slower than them, it can be very hard to compete because your competitor is launching features very quickly. But now that you have moved your China platform to the global platform, then you are prioritizing all those China features against other features in Germany, France and the UK. Indeed it was very hard.
Yes. The importance of localization can't be overemphasized and this is relevant for companies as well as for venture capital funds. It is the norm at this point.
Peter: Yes. Also the culture is very different. For example, at eBay the prevailing thought was that people are basically good, so when you buy a product on eBay, you pretty much just pay first, expecting the product to show up at your door a few days later. You don't even think about that. You just pay right away. But that trust was not there for China, so buyers didn't trust that the seller would ship the product.
Alibaba back then created an escrow mechanism whereby when you buy the product, you do not pay the seller. You pay to Alibaba first. They hold the money until the buyer received the product as advertised. Then you tell Alipay that I received the product and Alibaba will then release that money to the seller. So the culture is different.
Launching Eminence Ventures
How did you start investing in start-ups, and how did that lead to Eminence Ventures?
Peter: When I first got into VC I joined a boutique fund based out of Taipei. I was the only Partner based in China primarily responsible for internet investments. Back then I had a mentor, Jason Green, the Founding Managing Partner of Emergence Capital based out of Silicon valley.
Jason was an angel investor in AdChina, the company I co-founded, so when I got into VC he was one of the few people I knew that had been a very successful venture capitalist. Every time I went back to the States I would make an appointment with him and talk about VC and the challenges I encountered. Slowly he became my mentor.
Towards the end of 2016, my partner Simon and I decided we should probably be launching a very focused fund because the boutique VC fund that we were in was investing across all the sectors. We wanted our fund to be more focused. So Simon and I decided to launch a new fund focusing on enterprise cloud. We got that inspiration from Emergence as well.
Emergence has an incredible track record investing in general, but especially in the cloud area where some of their hits include Salesforce, Zoom (as the Series C lead), and more. For Eminence Ventures Fund I, how did you go about launching it? How were you able to raise capital and what did you invest in?
Peter: It was really hard. Coming from a non-VC background, as you could imagine, it was very hard to get institutional investors to invest in our first fund. We went after our network, our friends and friends of friends, and we were quite lucky that there were about 80 individuals who believed in us and they became our LPs in our fund one. That's how we got started.
And how big was the first fund that you raised?
Peter: We raised about $37 million for our fund I across two currencies, about 60% in the USD fund and 40% in the RMB fund.
Eminence Ventures Investment Strategy
That's pretty big for a first time effort. $37 million dollars is no chump change. Can you tell us a bit more about the fund’s strategy? What were you investing in when it comes to enterprise software?
Peter: If you look at our background, I had built a financial application, CRM application, infrastructure software, martech and an advertising platform, which were mostly in the B2B enterprise space. So naturally we gravitated towards that direction, given our computer engineering, product and sales and operating backgrounds. Throughout our careers, we also got to know a lot of people at pretty much all of the big brands, so we had a very extensive network there.
We thought enterprise cloud could be a natural sector for us. There are three areas we focus on: from the very top there is horizontal SaaS, ERP, CRM, human capital management, productivity and collaboration tools, etc. Secondly, there is vertical SaaS, industry focused cloud. And thirdly, at the very button is infrastructure, for example, data storage, database warehouse, security, dev ops tools, etc. These are the three main areas we look at. Obviously artificial intelligence is very important, but it can be applied across all the different applications or infrastructure software.
We invest across the stack, and are purely focused on the China market.
Understood. Could you share with us some examples of portfolio companies that you invested into and why?
Peter: Absolutely. We had our first close in August 2017 and wanted to be focused, so we started investing in a few areas that we were familiar with, of which one was obviously CRM.
One of the companies we invested in is called Recurrent.ai. They use natural language processing and AI to understand and process the conversations between telesales people and prospects. Recurrent technology analyzes these conversations and then provides feedback to telesales, empowering them and helping them become super sales.
In the States, there are a few comparable companies. For example, Gong.io (backed by Sequoia) and Chorus.ai and ASAPP, both of which were funded by Emergence who are in the similar space as Recurrent.ai.
Another example would be e-signature. In the US, you pretty much don't sign paper anymore these days. Back in 2017, the penetration rate of e-signature in China was extremely low, probably 1% to 3% give or take. We saw the opportunity and DocuSign didn't go public until 2018, so we thought it would be great timing to invest in an e-signature company. We decided to invest in Tsign back then and luckily it is becoming our first unicorn for Eminence. The company has grown a lot in the past three plus years.
Thanks for sharing those two examples. Digging a bit deeper on Tsign, as the market becomes more sophisticated and there's a shift towards digital signatures, I can definitely see this being increasingly a larger opportunity. I recall a few years back there were a few other companies in this space that raised fairly large amounts of funding including 法大大 (Fadada) and 上上签 (BestSign). Are there any other key players and how are they positioned against each other?
Peter: When we looked at the sector, there were over 30 players and we spoke to about 15 of those, and the few that you mentioned were actually the industry leaders. At that time, those two players were better funded, but we looked at the sector and determined that product and technology might not be the success factor. It’s just a signature, so if Baidu, Alibaba or Tencent wanted to launch an e-signature service it should not be too hard, it’s more or less a commodity. So if the product technology is not the determining success factor, what is? At that time, we came to the conclusion that one of them is going to be go-to-market strategy.
We liked the CEO Mr. Jin (金宏洲) who used to work at Alibaba in the so-called Iron Army as a salesperson. In the early days it meant a lot to work at the Iron Army team. For example, the CEO of DiDi (Cheng Wei) used to be part of that team as well. We liked Tsign’s go-to market strategy and the founder’s background. In addition, we figured that a big user of e-signature is going to be the government. Tsign is a very old company founded back in 2002 that started as a software service provider for many government agencies, so over the years they have accumulated all of the qualifications and credentials required to bid on government contracts. These two factors are just examples, but they helped us decide to invest in the company.
You mentioned that Tsign is worth about $1 billion, right? As we speak, DocuSign is worth about $38 billion in the US. As the China market continues to develop, how big can Tsign grow to and what might explain such a large valuation gap between players in the two markets.
Peter: I think the penetration of e-signature is still quite low in China today. I still remember when Alibaba founder Jack Ma said “you see it because you believe it.” So the question is, do you believe that in 3, 5, 7 years down the road, would e-signature become very popular in China? And I think we have that conviction. It will take a few years and we don't know when it will reach penetration rate of over 50 or even 70%, but I think it's going to come judging from the policies that we have seen issued by the government It’s moving towards that direction.
I won't be surprised if they are multiple unicorns born out of this sector. I think Tsign should have a long way to go, so we are holding onto our stake. After our investment, Alibaba's ANT Financial led the following round and we also double downed on that round.
Overall we are pretty optimistic about Tsign and the entire sector.
Well that sounds like a solid development for the company. As you described, the usage is growing but it's still fairly early. Something I tend to think about when I look at enterprise software and cloud deals in China is that… it seems the willingness to pay by Chinese companies for some of the cloud/software products sometimes isn’t there.
How do you think about this, and what kind of products and software do people want to pay for? If we take Recurrent.ai as an example, its model generally makes sense to me, since the client could generate higher sales using the product. Recurrent.ai software literally helps the client convert more deals and leads.
Peter: I think it's important and we’ve invested in quite a few so-called Wechat CRM companies that help enterprises grow their customer base and acquire and convert new users.
We need to look at the US market as well. For example, there's a VC firm called Bessemer which has been tracking so-called emerging cloud companies. I study their State of the Cloud report every year. If we go back to 2010, in the US there was only one cloud unicorn according to Bessemer, but last year if you look at their Cloud 100 list, you won’t even make it to the list if you are not a unicorn. So 10 years is a huge difference even in the US.
If you look at some of the cloud companies that have gone public, for example, Kingsoft Office is about $25 billion market cap and Agora used to be $10 billion market cap, which has probably gone down to $5 billion, but still not small. And if we were to look at when Salesforce went public, Salesforce was a $1.1 billion company. When Shopify went public, it was maybe a $3 to $4 billion company. They were small at IPO, but they grew over the years. That's the beauty of the SaaS model. So I think we need to look at the cloud history in the US and then look at China. If you look ahead to the next 10 years, there will be tremendous growth.
Another example is Ming Yuan Cloud, which is a Hong Kong listed company with a market cap of about 80 billion HKD, so over 10 billion USD. They are in the real estate SaaS business providing ERP and CRM applications just for real estate developers, and the company is trading at more than 40x multiple. There are quite a few examples with nice valuations and for the ones that we talked about, they all have strong revenues.
We also have been studying some of the China SaaS reports from leading investment banks such as JP Morgan, Citibank and Credit Suisse. Recently, according to JP Morgan, they predicted that by 2025, the market cap of China SaaS companies will reach close to $1 trillion USD. This is not our prediction but JP Morgan’s, so I think maybe JP Morgan is even more optimistic than us.
Key Success Factors for Building a Software Company in China
From the company's perspective, what should they be thinking about? What are the key success factors for them?
Peter: If you want to build an enterprise software, you need to find a true customer need and solve their pain points. You can provide value by increasing productivity, increasing sales or reducing cost, and delivering a superior product. We always tell entrepreneurs from day one that if you want to start your business, you need to think about if Alibaba, Tencent, ByteDance or Baidu were to build a similar product, how you are going to defend yourself. That’s a very natural question. So obviously you have to have a superior product, you have to move faster and you need to have unique go-to market strategies that work, and lastly, you have to attract and retain talents. These are all the key success factors.
At a high level, when these companies have the product ready, how should they sell, and who should they sell to? Should these be SMBs, internet companies, large enterprises or state-owned enterprises? What should they keep in mind?
Peter: I think there are quite a few opportunities across various sectors you can go after, for example the so-called key accounts, the big enterprises. That's what Recurrent.ai is going after, the largest insurance companies, the largest banks, the largest online education companies.
Those are the companies with the most telesales people. You can also go after e-commerce sellers because so many of them have been educated by Alibaba, JD and PDD to pay for their services, so it's a very good sector. Most of them are SMBs because there are so many small sellers on those platforms, so there are a lot of opportunities. However, if you are going after the regular SMBs who are not really in the e-commerce sector, I think we need to have a little bit more patience because their willingness and ability to pay in the next year or two may still be a little lacking, vs if you go after the mid-tier. Mid tier could be a good spot because they don't require a lot of customization and they also have the scale and the ability and willingness to pay.
You mentioned that certain types of clients might have lower willingness to pay in the near term. What are some other barriers to adoption? For example, level of IT maturity, etc. What are the requirements when it comes to customization?
Peter: A lot of large enterprises have special needs that require full customization, so it’s very important to keep this in mind when you build a software. You need to have this Platform-as-a-Service layer that allows the large enterprises to build their own functionality. If you don't have that PaaS layer, then you could be in the business of doing professional services or customization which is not really scalable.
For SMBs, I would say I'm more optimistic even though for the time being their ability and willingness to pay is not quite like that in the States, but they are growing very fast. If you look at the key reason that US enterprises are so willing to pay for software is because people are expensive and the labor costs are so high, so anything you can do to increase the productivity of their employees, they're willing to pay for it. The pay scale in China is pretty high now too, particularly for the professional workers, great software engineers, great designers, great marketers, etc. Those people are very expensive, so I think the willingness and ability to pay will come very quickly in the next couple of years.
Yes. I definitely see that happening as part of a greater macro trend. China has seen incredible growth in the past 10 years when it comes to tech startups and the venture capital industry. A lot of the fast and higher returning deals tended to come from the consumer internet category. But now as GDP per capita hits $10k USD, the cost is going up and there's more of a need to focus on efficiency and productivity, in reducing costs or in some cases augmenting sales capabilities. That makes a lot of sense.
Peter: Also giants like Alibaba and Tencent have cloud services so they are actually educating the market and encouraging companies to move everything to the cloud. Ali Cloud and Tencent Cloud are growing very fast.
On the topic of cloud. How do Chinese companies choose between public vs. private cloud setup?
Peter: Ali Cloud, Tencent Cloud and Huawei Cloud, as examples, all have public cloud service so you can just go and open up an account and subscribe to them. They can give you 100 virtual machines in a few minutes and then you can start launching your applications. So public cloud is a no brainer if you can just use them, but some of the enterprises, particularly bigger ones or state-owned enterprises, may be concerned about data security so they may not be willing to put their data or applications on the public cloud. In those cases, they may be renting a server or a virtual machine from the cloud provider, but this virtual machine or this server is dedicated to them and no other clients who will be sharing the space.
There are some state-owned enterprises or big enterprises who prefer to do it that way. Obviously it might be more expensive, but these are the bullets you have to bite. But if they want the software to be installed on-premises, then it will be more challenging because it's going to be similar to the model when I worked at Oracle back in 1997 to 2000, which is impossible to maintain. We don't like that model. There are some software startups who started to entertain those businesses, and those are not very scalable businesses so we try to stay away from them.
I see. Yes, on-prem would be pretty difficult, but when it comes to private versus public cloud, what is the business model there? How do these companies charge for their solutions? Is it buying a SaaS model or is software licensing plus service?
Peter: I think most of the mid-tier companies and SMBs are okay with the subscription model. Normally they just pay on an annual basis like the Zoom subscription model. But some large enterprises may not be willing to pay on an annual basis, partially because their budget is not being set up that way, particularly the state-owned enterprises. In that case they can only purchase the license. Instead of say $100k, they may be willing to pay you $300k of $400k upfront, and then after that, every year you charge them somewhere between 10% to 20% for service charge. Overall I think more and more companies are willing to pay on a subscription basis, which is a good sign.
What was the impact of COVID and how did it affect development of this sector in China?
Peter: Similar to the States, COVID has had a very positive impact on SaaS and cloud companies. For example, we would not imagine that the ministry of housing in China and also a couple of other ministries would issue policies encouraging companies to sign contracts with their employees digitally. And now if you want to rent an apartment or buy a house, the government encourages you to sign the contract digitally.
Due to COVID, this digital transformations has become much faster. Another example is Recurrent.ai that we talked about. During COVID people started to know that they could get sick, so they wanted to buy insurance but didn’t want to meet with insurance agents. Online meetings become the norm and there are a lot of telesales conversations that need to be facilitated and tracked, and sales people need to be empowered. COVID really helped a lot with software and when you look at enterprise software companies that are listed, their stock prices have gone through the roof in the past 18 months.
Yes. I definitely see that in the US. Just to clarify on Recurrent.ai. Do they record all types of audio, video calls, etc.? Conceivably, they could leverage their tech to provide services in areas beyond just improving sales conversion? Are you thinking about other directions for the technology as well?
Peter: There are quite a few other applications. They don't record the conversations by the way. It’s their clients who are recording those conversations. For example, in the US, you know when you get into a conversation, they will prompt you saying that this conversation will be recorded for training purposes. So those enterprises are already recording the conversations for their own training or quality assurance purposes, and Recurrent.ai is just scanning those recording from their clients and pretty much helping their clients with understanding the customer prospects and training the telesales people with responses that will give them a higher probability of closing the deal.
Education, insurance and banking are the main sectors. Real estate, car dealerships and even consumer are important sectors - imagine P&G and Unilever each have about 10,000 promoters working at Walmart, Carrefour, etc. Those promoters have conversations with consumers on a daily basis. These are the cases where you have salespeople interacting with prospects.
Definitely. I was thinking even like concierge services at a hotel or front desk and how best to service some of the clients. Some of that is sales, but some of that is also customer service and improving the end user experience or the customer experience. I think that's really cool.
Peter: Yes. For Recurrent.ai they want to first go after the sectors whose clients are the most willing to pay, and those would be education, insurance and banks. Then can expand around that.
Peter, I'm conscious of your time as it's getting pretty late. I wanted to ask you one more thing, which is what’s next for you? You just raised Eminence Venture Fund II at quite a large fund size of about $120 million. How were you thinking about deploying the fund and what are you particularly excited about over the next 5-10 years?
Peter: We have made three investments so far. We will probably have about 20 companies in our fund II portfolio and so we are still in the early phase of our capital deployment. On some sectors we are pretty excited about, one is the WeCom (formerly WeChat Work) ecosystem, and another one is what we call deep collaboration. This last piece was inspired by Emergence Capital. If you go to their website, one of their GPs Jake wrote an article about deep collaboration.
This is another area we are excited about. In fact, the three companies that we have invested so far are all in deep collaboration. Industry cloud and infrastructure software related to data are something we are looking into, because as you can imagine, there will be so much data being accumulated, and there will be a lot of tools and applications that are needed for handling, processing and doing all kinds of things related to data.
Yes, it sounds like there's plenty for you and your team to sink your teeth into. Best of luck to you, Peter, and to your team. Hopefully you can invest in and help build more cloud unicorns in China.
Peter: Thank you so much for your time and for the interview as well. Hope you have a good night.